How to Improve Vocabulary — Methods That Hold Up

Recognition and production are separate piles

You can read a word, understand it perfectly in context, and be entirely unable to summon it when you need it. That is not forgetting — the two abilities are stored and retrieved differently, and only one of them was ever practiced.

The consequence is practical. If you want a word available when you speak, you have to speak it. If you want it available when you write, write it. The route into production is production, and there is no reading-only path.

The cheapest version is a use-it-once rule: when a word interests you, put it into a real sentence about your own life that same day. Not a textbook sentence — one with your actual Tuesday in it. The specificity is what gives the word somewhere to attach.

Our vocabulary test measures the recognition pile, which is the useful starting number because it tells you whether the gap is knowledge or retrieval. Someone who recognizes a great deal and produces little has a different problem from someone who genuinely has not met the words.

Context beats lists, for a reason

Words learned inside a sentence outlast words learned as definitions, because a definition gives you one retrieval route and a sentence gives you several.

A word met in a story carries the situation, the tone, the register, and often who would say it — and any of those can pull it back later. A word met on a list carries only the definition, so if the definition does not surface, nothing does. This is also why list-learned words come out wrong in a way that is hard to explain: you learned what it means without learning where it belongs.

The practical rule is to always keep one real example. When you note a word, note the sentence it appeared in, not just the meaning. It costs a few extra seconds and roughly doubles what you can do with the word afterwards.

Wordplay does the same job in a more entertaining form, which is why it is worth using deliberately. Our riddles work on double meanings, our idioms quiz works on fixed expressions where the meaning is not the sum of the parts, and our slang quiz covers the layer that changes fastest and is rarely taught anywhere.

Choosing which words are worth the effort

Not all words repay the same effort, and a good filter saves more time than a good method. There is a number attached to that filter. Nation (2006) found that reading a wide range of authentic texts without unknown words getting in the way takes roughly 8,000 to 9,000 word families, and listening takes 6,000 to 7,000. The words worth chasing sit inside that band, not out in the rare ones.

The words worth deliberate work share three properties: you have met them more than once, they belong to something you talk or write about regularly, and they say something your current vocabulary cannot say in fewer words. That last test is the strongest one — a word that is merely a longer synonym for something you already have adds nothing.

What is usually not worth deliberate work is the rare and impressive-sounding. Those words are memorable precisely because they are rare, which also means the chance to use one arrives about twice a year, and using it wrongly is more noticeable than not using it at all.

The useful counterintuitive move is to spend effort on precision inside common ground rather than on exotic ground. Knowing four accurate words for degrees of annoyance is worth more in practice than knowing one spectacular word for a situation you will never be in.

A routine that survives a normal week

Ten minutes most days beats an hour on Sunday, and the reason is retrieval spacing rather than total time.

A version that holds up:

1. Collect during the day. One note, the word plus the sentence you met it in. No editing.
2. Once a day, take three from the collection and use each in a sentence about your own life, out loud or written.
3. Once a week, go back three weeks and try to produce the meanings from memory before checking.
4. Drop anything you have failed twice and never had a use for. It is not a moral failing; it is a bad candidate.

Step four is the one people resist and it matters. A collection you never prune becomes a list you avoid, and an avoided list teaches nothing. Cutting the words you have no occasion to use makes the rest more likely to actually get practiced.

If the goal is a second language rather than a first, the same routine applies with one addition: check where the word sits in the level system before deciding it is essential. Our guide on CEFR levels covers that, and it prevents a common failure where someone drills advanced vocabulary while their sentence structure is two levels behind.

What does not work as advertised

A few widely recommended methods produce a feeling of progress out of proportion to the result.

Word-a-day services are the clearest case. A word arrives with no connection to anything you were doing, gets read, and disappears, because nothing in your week gave it a reason to come back. The service is not the problem — the missing use is.

Memorizing synonym lists has a similar issue and adds a specific hazard. Words presented as synonyms are rarely interchangeable in register or connotation, so learning them as equivalents produces sentences that are technically correct and audibly wrong.

And reading with a dictionary open for every unknown word tends to destroy the reading without helping the vocabulary. A better rule is to keep reading through the first two unfamiliar words and only stop for the third, or for one that clearly carries the sentence. Uninterrupted reading builds the contextual sense that makes the next unknown word guessable, which is the ability that scales.

If your aim is broad knowledge rather than expression, that is a different target with a different method, and our guide on getting better at general knowledge covers it.